Genome-wide analysis of hepatic DNA methylation reveals impact of epigenetic aging on xenobiotic metabolism and transport genes in an aged mouse model

Abstract

Hepatic xenobiotic metabolism and transport decline with age, while intact xenobiotic metabolism is associated with longevity. However, few studies have examined the genome-wide impact of epigenetic aging on these processes. We used reduced representation bisulfite sequencing (RRBS) to map DNA methylation changes in liver DNA from mice ages 4 and 24 months. We identified several thousand age-associated differentially methylated sites (a-DMS), many of which overlapped genes encoding Phase I and Phase II drug metabolizing enzymes, in addition to ABC and SLC classes of transporters. Notable genes harboring a-DMS were Cyp1a2, Cyp2d9, and Abcc2 that encode orthologs of the human drug metabolizing enzymes CYP1A2 and CYP2D6, and the multidrug resistance protein 2 (MRP2) transporter. Cyp2d9 hypermethylation with age was significantly associated with reduced gene expression, while Abcc2 expression was unchanged with age. Cyp1a2 lost methylation with age while, counterintuitively, its expression also reduced with age. We hypothesized that age-related dysregulation of the hepatic transcriptional machinery caused down-regulation of genes despite age-related hypomethylation. Bioinformatic analysis of hypomethylated a-DMS in our sample found them to be highly enriched for hepatic nuclear factor 4 alpha (HNF4α) binding sites. HNF4α promotes Cyp1a2 expression and is downregulated with age, which could explain the reduction in Cyp1a2 expression. Overall, our study supports the broad impact of epigenetic aging on xenobiotic metabolism and transport. Future work should evaluate the interplay between hepatic nuclear receptor function and epigenetic aging. These results may have implications for studies of longevity and healthy aging.

Keywords: Aging; Cytochrome P450s; Epigenetics; Hepatic transporters; Hepatic xenobiotic metabolism.

Fig. 1

A Plot of sample loadings on the first two principal components of the RRBS DNA methylation data and B Clustering of samples, based on correlation and the Ward clustering method (analysis from MethyKit)

Fig. 2

Volcano plot of age-associated differentially methylated sites in mouse liver from age 4 to 24 months. The y-axis is the negative logarithm of the p-value, while the x-axis shows the methylation change with age. Genome-wide significance is determined by false discovery rate (FDR) less than 5% (horizontal dotted line). Vertical dotted lines show > 10% hypermethylation and > 10% hypomethylation with age. Genome-wide significant loci showing > 10% hypermethylation with age are color coded red, while > 10% hypomethylation are blue. The genes associated with the top 5 hyper- and hypomethylated loci are shown, as are top xenobiotic metabolism genes and transporters

Fig. 3

Gene ontology (GO) enrichment analysis of genes with age-associated differentially methylated sites > 10% hyper- or hypomethylated in our primary analysis. The most enriched biological processes are shown ordered by decreasing statistical significance. Analysis and plot from GREAT

Fig. 4

DNA methylation and expression changes with age for Cyp2d9, Cyp1a2, and Abcc2. Panels A), D) and G) show the most significant change in DNA methylation with age at the three loci, with Cyp2d9 showing hypermethylation with age, while Cyp1a2 and Abcc2 show hypomethylation with age (data from RRBS). Panels B), E) and H) show the relative quantification of RNA change with age, where 4-month-old samples are set mean = 1.0. Both Cyp genes showed lower gene expression with age while there was no significance difference in Abcc2 expression with age. Panels C), F) and I) show the relationship between methylation levels and expression of the gene. For Cyp2d9 there was a significant negative relationship between the two variables, whereas for Cyp1a2 and Abcc2 there was no significant association between DNA methylation change with age and expression. (n = 6 mice per age group)

Fig. 5

Replication of DNA methylation and expression changes with age for Cyp2d9 and Cyp1a2. Panels A) and D) show the change in DNA methylation with age at the most significant a-DMS at Cyp2d9 and Cyp1a2 respectively, in an independent sample of male subjects (n = 6 for each age group of 4 and 24 months old) using a targeted technology, high resolution melt (HRM) analysis of bisulfite converted DNA. Panels B) and E show the change in gene expression with age for Cyp2d9 and Cyp1a2 respectively, in the replication sample. Panel C) shows a significant negative relationship between Cyp2d9 DNA methylation and expression, whereby increased DNA methylation leads to reduced expression, in the combined discovery and replication samples (n = 12 for each age group of 4 and 24 months old). Panel F) shows the lack of relationship between methylation and expression of Cyp1a2 in the combined discovery and replication samples